Comparative Analysis of Novel View Synthesis and Photogrammetry for 3D Forest Stand Reconstruction and Extraction of Individual Tree Parameters

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(9), P. 1520 - 1520

Published: April 25, 2025

The accurate and efficient 3D reconstruction of trees is beneficial for urban forest resource assessment management. Close-range photogrammetry (CRP) widely used in the model scenes. However, practical forestry applications, challenges such as low efficiency poor quality persist. Recently, novel view synthesis (NVS) technology, neural radiance fields (NeRF) Gaussian splatting (3DGS), has shown great potential plants using some limited number images. existing research typically focuses on small orchards or individual trees. It remains uncertain whether this technology can be effectively applied larger, more complex stands In study, we collected sequential images plots with varying levels complexity imaging devices different resolutions (cameras smartphones UAV). These included one sparse, leafless another dense foliage occlusions. We then performed NeRF 3DGS methods. resulting point cloud models were compared those obtained through photogrammetric laser scanning results show that to method, NVS methods have a significant advantage efficiency. method suitable relatively simple stands, it less adaptable ones. This tree issues excessive canopy noise wrongfully reconstructed duplicated trunks canopies. contrast, better adapted yielding clouds highest offer detailed trunk information. lead errors ground area when input views are limited. capability generate clouds, density, particularly sparse points areas, which affects accuracy diameter at breast height (DBH) estimation. Tree crown information extracted from by all three methods, achieving height. DBH still higher than clouds. Meanwhile, ground-level smartphone images, parameters higher-resolution varied perspectives drone accurate. findings confirm application forests.

Language: Английский

---

Simplifying drone-based aboveground carbon density measurements to support community forestry

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(4), P. e0322099 - e0322099

Published: April 29, 2025

Community-based forest restoration has the potential to sequester large amounts of atmospheric carbon, avoid degradation, and support sustainable development. However, if partnered with international funders, such projects often require robust transparent aboveground carbon measurements secure payments, current monitoring approaches are not necessarily appropriate due costs, scale, complexity. The use consumer-grade drones in combination open source structure-from-motion photogrammetry may provide a solution. In this study, we tested suitability simplified drone-based method for measuring density heavily degraded tropical forests at 2 ha site Sabah, Malaysia, comparing our results against established field-based methods. We used generate canopy height models from drone imagery, applied multiple pre-published plot-aggregate allometric equations examine importance utilising regionally calibrated equations. Our suggest that can produce similar magnitude methods, quickly only single input metric. there greater levels uncertainty errors associated drones. findings also highlight selecting approach. At scales between 1 100 ha, methods an appealing option data acquisition measurement, balancing trade-offs accuracy, simplicity, cost effectiveness coinciding well needs community-scale measurement. Of importance, discuss considerations relating accessibility community use, beyond purchasing drone, must be overlooked. Nevertheless, presented here lays foundations simple workflow scale refined future studies.

Language: Английский

---

3D Reconstruction of a Complex Grid Structure Combining UAS Images and Deep Learning

Remote Sensing, Journal Year: 2020, Volume and Issue: 12(19), P. 3128 - 3128

Published: Sept. 23, 2020

The latest advances in technical characteristics of unmanned aerial systems (UAS) and their onboard sensors opened the way for smart flying vehicles exploiting new application areas allowing to perform missions seemed be impossible before. One these complicated tasks is 3D reconstruction monitoring large-size, complex, grid-like structures as radio or television towers. Although image-based survey contains a lot visual geometrical information useful making preliminary conclusions on construction health, standard photogrammetric processing fails dense robust complex large-size mesh structures. main problem such objects repeated self-occlusive similar elements resulting false feature matching. This paper presents method developed an accurate Multi-View Stereo (MVS) Shukhov Radio Tower Moscow (Russia) based UAS survey. A key element successful WireNetV2 neural network model automatic semantic segmentation wire proposed provides high matching quality due masking tower elements. contributions are: (1) deep learning convolutional that outperforms state-of-the-art results dataset containing images grid topology with elements, holes, self-occlusions, thus providing structure and, result, reconstruction, (2) advanced pipeline aided by structured, evaluated imagery Moscow.

Language: Английский

---

Assessment of the Influence of Survey Design and Processing Choices on the Accuracy of Tree Diameter at Breast Height (DBH) Measurements Using UAV-Based Photogrammetry

Drones, Journal Year: 2021, Volume and Issue: 5(2), P. 43 - 43

Published: May 24, 2021

This work provides a systematic evaluation of how survey design and computer processing choices (such as the software used or workflow/parameters chosen) influence unmanned aerial vehicle (UAV)-based photogrammetry retrieval tree diameter at breast height (DBH), an important 3D structural parameter in forest inventory biomass estimation. The study areas were agricultural field located province Málaga, Spain, where small group olive trees was chosen for UAV surveys, open woodland area outskirts Sofia, capital Bulgaria, 10 ha grove, composed mainly birch trees, overflown. A DJI Phantom 4 Pro quadcopter image acquisition. We applied structure from motion (SfM) to generate point clouds individual using Agisoft Pix4D packages. estimation DBH made RANSAC-based circle fitting tool TreeLS R package. All modeled had their tape-measured on ground accuracy assessment. In first site, we executed many diversely designed flights, identify which parameters (flying altitude, camera tilt, method) gave us most accurate estimations; then, resulting best settings configuration assess replicability method forested Bulgaria. tested (flight altitudes about 25 m above canopies, tilt 60°, forward side overlaps 90%, ultrahigh processing) resulted root mean square errors (RMSEs; %) below 5% diameters site 12.5% area. demonstrate that, when carefully methodologies are used, SfM can measure single with very good accuracy, our knowledge, results presented here achieved so far (above-canopy) UAV-based photogrammetry.

Language: Английский

---

Improving Urban Mapping Accuracy: Investigating the Role of Data Acquisition Methods and SfM Processing Modes in UAS-Based Survey Through Explainable AI Metrics

Journal of Geovisualization and Spatial Analysis, Journal Year: 2024, Volume and Issue: 8(1)

Published: May 9, 2024

Abstract In this study, we investigated the accuracy of surface models and orthophoto mosaics generated from images acquired using different data acquisition methods at processing levels in two urban study areas with characteristics. Experimental investigations employed single- double-grid flight directions nadir tilted (60°) camera angles, alongside Perimeter 3D method. Three (low, medium, high) were applied SfM software, resulting 42 models. Ground truth RTK GNSS points aerial LiDAR surveys used to assess horizontal vertical accuracies. For test, neither oblique angle nor double grid resulted an improvement accuracy. contrast, when examining accuracy, it was concluded that for several levels, yielded better results, these cases, also improved Feature importance analysis revealed that, among four variables, method most important factor affecting out three cases.

Language: Английский

---

Measures of Canopy Structure from Low-Cost UAS for Monitoring Crop Nutrient Status

Drones, Journal Year: 2020, Volume and Issue: 4(3), P. 36 - 36

Published: July 22, 2020

Deriving crop information from remotely sensed data is an important strategy for precision agriculture. Small unmanned aerial systems (UAS) have emerged in recent years as a versatile remote sensing tool that can provide precisely-timed, fine-grained informing management responses to intra-field variability (e.g., nutrient status and pest damage). UAS sensors with high spectral resolution used compute informative vegetation indices, however, are practically limited by cost dimensionality. This research extends analysis monitoring investigate the relationship between health 3D canopy structure using low-cost equipped consumer-grade RGB cameras. We flue-cured tobacco case study due its known sensitivity fertility variation nutrient-specific symptomology. Fertilizer treatments were applied induce plant 0.5 ha field of tobacco. Multi-view stereo images three surveys collected during development processed into orthoimages visible band index photogrammetric point clouds Structure Motion (SfM). Plant structural metrics then computed detailed surface models (0.05 m resolution) interpolated clouds. The complimented measurements obtained tissues. relationships foliar nitrogen (N), phosphorus (P), potassium (K), boron (B) concentrations UAS-derived assessed multiple linear regression. Symptoms N K deficiencies well captured differentiated metrics. strongest observed was shape concentration (adj. r2 = 0.59, increasing adj. 0.81 when combined index). B consistently better predicted maximum 0.41 at latest growth stage surveyed. Overall, combining about reflectance increased model fit all measured nutrients compared alone. These results suggest exists relative be leveraged improve usefulness low

Language: Английский

---

An assessment of conventional and drone-based measurements for tree attributes in timber volume estimation: A case study on stone pine plantation

Ecological Informatics, Journal Year: 2021, Volume and Issue: 63, P. 101303 - 101303

Published: April 15, 2021

Language: Английский

---

Mapping Canopy Heights in Dense Tropical Forests Using Low-Cost UAV-Derived Photogrammetric Point Clouds and Machine Learning Approaches

Remote Sensing, Journal Year: 2021, Volume and Issue: 13(18), P. 3777 - 3777

Published: Sept. 20, 2021

Tropical forests are a key component of the global carbon cycle and climate change mitigation. Field- or LiDAR-based approaches enable reliable measurements structure above-ground biomass (AGB) tropical forests. Data derived from digital aerial photogrammetry (DAP) on unmanned vehicle (UAV) platform offer several advantages over field- in terms scale efficiency, DAP has been presented as viable economical alternative boreal deciduous However, detecting with ground dense forests, which is required for estimation canopy height, currently considered highly challenging. To address this issue, we present generally applicable method that based machine learning methods to identify forest floor DAP-derived point clouds We capitalize high-resolution vertical inform detection. conducted UAV-DAP surveys combined field inventories Congo Basin. Using airborne LiDAR (ALS) truthing, height model (CHM) generation workflow constitutes detection, classification interpolation points using combination local minima filters, supervised algorithms TIN densification classifying spectral geometrical features UAV-based 3D data. demonstrate our DAP-based provides estimates tree heights identical (conservatively estimated NSE = 0.88, RMSE 1.6 m). An external validation shows capable providing accurate precise AGB (DAP vs. old forest: r2 0.913, 31.93 Mg ha−1). Overall, study demonstrates application cheap easily deployable platforms can be deployed without expert knowledge generate biophysical information advance monitoring

Language: Английский

---

Earthwork Volume Calculation, 3D Model Generation, and Comparative Evaluation Using Vertical and High-Oblique Images Acquired by Unmanned Aerial Vehicles

Aerospace, Journal Year: 2022, Volume and Issue: 9(10), P. 606 - 606

Published: Oct. 15, 2022

In civil engineering and building construction, the earthwork volume calculation is one of most important factors in design construction stages; therefore, an accurate necessary. Moreover, because managing earthworks highly important, this study, a three-dimensional (3D) model for management was performed using unmanned aerial vehicle (UAV) RGB camera. Vertical high-oblique images (45°, 60°, 75°) were acquired at 50 100 m heights calculations 3D model, data generated by dividing into eight cases. Cases 1–4 from height m, cases 5–8 m. (case 1: 90°, case 2: 90° + 45°, 3: 4: 75°, 5: 6: 7: 8: 75°). Three evaluations on data. First, accuracy evaluated through checkpoints orthophoto; second, volumes calculated via global positioning system UAV compared; finally, evaluated. Case 2, which showed lowest root mean square error orthophoto evaluation, accurate. 2 evaluation compared to other Through best results obtained when vertical image 40 50° generating management. addition, if not affected obstacles, it better shoot about or less than too high.

Language: Английский

---

Method of 3D Voxel Prescription Map Construction in Digital Orchard Management Based on LiDAR-RTK Boarded on a UGV

Drones, Journal Year: 2023, Volume and Issue: 7(4), P. 242 - 242

Published: March 30, 2023

Precision application of pesticides based on tree canopy characteristics such as height is more environmentally friendly and healthier for humans. Offline prescription maps can be used to achieve precise pesticide at low cost. To obtain a complete point cloud with detailed information in orchards, LiDAR-RTK fusion acquisition system was developed an all-terrain vehicle (ATV) autonomous driving system. The transformed into geographic coordinate registration, the Random sample consensus (RANSAC) segment it ground canopy. A 3D voxel map unit size 0.25 m constructed from cloud. 20 trees geometrically measured evaluate accuracy map. results showed that RMSE between calculated LiDAR obtained actual 0.42 m, relative (rRMSE) 10.86%, mean absolute percentage error (MAPE) 8.16%. generate meet requirements precision application. could construct autonomously match digital orchard management.

Language: Английский

---